EFEKTIVITAS PENERAPAN MODEL PEMBELAJARAN GENERATIF UNTUK MENGATASI MISKONSEPSI PADA MATERI SEGIEMPAT SISWA SEKOLAH DASAR

This study aims to identify the types of misconceptions in quadrilateral topics among elementary school students, the causes of these misconceptions, and the effectiveness of the generative learning model in addressing misconceptions in quadrilaterals among elementary school students. The research is a quasi-experimental study without a control group, in which a single treatment the application of the generative learning model is administered to one group. The method used is a mixed-method research approach with a concurrent embedded strategy. The population consists of Grade VI students at Muhammadiyah 2 Elementary School in Tarakan during the first semester of the 2024/2025 academic year, comprising three parallel classes. The subjects are 25 students from the Al Kautsar class, selected through cluster random sampling. Data were collected using research instruments such as learning achievement tests, interviews, observation sheets, and questionnaires. The results of the study are as follows: (1) Elementary school students experience four types of misconceptions in quadrilateral topics: pre-conception, undergeneralization, prototyping error, and overgeneralization; (2) The causes of misconceptions in quadrilateral topics among elementary school students include six factors: (a) Limited understanding of definitions and properties, (b) Visualization with less variety in forms, (c) Influence of incorrect information or explanations, (d) Monotonous teaching patterns or schemas, (e) Insufficient variation in examples provided, and (f) Lack of understanding of shape hierarchies; (3) The generative learning model is effective in addressing misconceptions in quadrilateral topics among elementary school students. This is evidenced by: (a) The average post-test scores of Grade VI students at Muhammadiyah 2 Elementary School in Tarakan reached the Minimum Competency Criteria (KKM) with an average score of 84.56, achieving classical completeness at 88.00%, which exceeds 85%. The average normalized gain score was 0.6672, categorized as " medium," and statistical hypothesis testing indicated that the null hypothesis (H_0) was rejected, (b) The average percentage of student activities aligned with the learning process was 61.33%, categorized as "good.", (c) The average percentage of teachers' ability to manage learning using the generative learning model was 95%, categorized as "excellent”, (d) The average percentage of students who responded positively to learning using the generatif learning model was 85.3%, exceeding 75%.